TW565810B - Three-dimensional electronic map information generation method - Google Patents

Abstract

Make 3D modeling of building highly efficient. When generating 3D electronic map information, use the following procedures that won't measure in terms of building's height to conduct 3D modeling of building. Firstly, after taking picture of building, record the shooting position and shooting parameters (camera's orientation, image angle). Secondly, arrange pictures in accordance with these information to revive the state when taking the picture in the virtual space prepared on computer. Thirdly, Move surface shape along attitude direction to overlap with picture for building modeling. Thus doing so without the measurement of actual building along the attitude can finish 3D modeling.

Description

565810

FIELD OF THE INVENTION The present invention relates to a technology for generating a building meta-modeling technology and a technology for using the modeling material. Three-dimensional method of three-dimensional electronic data of materials. Three-dimensional electronic map information learning technology has been widely used in recent years (hereinafter referred to as "electronic map data") in personal computer map display, road map supply and printing In the navigation system, the use of the three-way forward route is gradually used. By grasping the three-dimensional and forward route, it becomes easier. In order to display in two-dimensional, it requires a large amount of labor technology for the information about the height of the map object. The technology recorded in this technology, the photogrammetry and remote control "About the three-dimensional simulation based on the lens image" by the University of Tokyo Institute of Production and Measurement, etc. The former is to reduce the distance measurement result and C of the data distance device by and efforts. c D 影 {type. Electronic map data available to the brain) ° Electronic map data is used for so-called navigation systems for vehicles, and the production of maps via Internet products. In addition, the 疋 display allows the driver to intuitively display buildings and the like, which has the advantage of present position. Three-dimensional information on buildings is required. Obtain three-dimensional data of many buildings, especially labor. Various proposals have been made in the past to alleviate this. For example, there are patents No. 3, No. 3, No. 3, No. 3, No. 3, Zhao Huijing, Japanese Writer Sense, Vol. 36, No. 4, 1997, Automatic Measurement of Yuancheng Spatial Data. Labor required for measuring the height of a building. The latter is a method of obtaining three-dimensional data by overlapping laser gauges.

Bing 5810 V. Description of the invention (2) Modeling, 21 = Γ The method of obtaining the three-dimensional data of the soil to achieve the order of the three-dimensional building "degree-independent assumption of a fixed height or self-virtual = two degrees method ... are not fully reflective of the reality, < a large 7L modelling method. Modeling the state of the problem to be solved by the invention: f-type method Some kind of measurement of each building. In order to produce a product :::: Data 'needs to be used for a large number of buildings: reason:' the conventional technology is used to measure the reduction of labor. The electronic map data of Ren 70 is also very demanding. The unit has been newly built and demolished for several years. Because of the practical use of this map data, it can be traced to the three needs; meta-modeling. Known technology The law also fully meets this period.

In view of the above-mentioned lesson, the present invention makes it possible to reflect a variety of techniques with light labor, and the model of lightness, lightness, and reusability is 70 times. And its purpose. State L’ s use of this three-dimensional model to produce products in a short period of time ... Resources: the use of methods to solve the problem and its effect • Effect In order to solve at least one problem of this problem, in the present invention, By using page 62061-4944-PF (N); ahddub.ptd 565810 V. Description of the invention (3) Comparison of the state of the computer and the like in the virtual prepared space, the amount k without accompanying height is realized Object t photo II, = structure of the invention, before the description of the invention; Represents: 1 of the electrical brain surface: the principle of the three-dimensional model of the invention. The situation of the virtual space which is difficult to be displayed by the monitor. The three-dimensional coordinates such as latitude A and 鬲 are used to define the virtual *. The initial stage 'is displayed on the display DISP port as Lu Jiao's office worker m on the surface of the virtual space. In the virtual space prepared by the ic, the plane shape ν of the three-dimensional model object is arranged. Yi Zhao's photo PIC and the subject To, silly smile two Photography position, the side of the camera Dry :! Configure photo PIC. The photo pic configured in this way is ί = ίΐ The image will be displayed on the object. In addition, the two-dimensional shape PV of the Thousands of Faces of objects can be easily followed by the two-dimensional map data that already exists. As shown in the lower part of the figure, the plane shape is generated in a virtual space in a highly square shape like this; = As shown above, the photo PIC In the virtual space, if a moving plane shape PV exists and the modulus direction is parallel to the shift, "-when the model and the image PB overlap, today-two people ^ become a reflection of the actual height of the object model. In = example, in order to / reproduce the scenery obtained by using photography like this, use the ancient: direction ί ι to move the plane shape 实现 to realize the three-dimensional model of the object: again in Figure 1 according to the computer display of the day to illustrate the invention The principle Ύ. Page 7 2061 -4944-PF (N); ahdd ub.p t d 565810 V. Description of the invention (4) The state at the time of explaining the principle of the present invention. Fig. 2 is an explanatory diagram showing the M-series of the image and the killer. The top view at the time of photographing is VI, the side view is at the center ^ 2, and the case of ± _PIC at the bottom.所 Photos not shown below

As shown in plan view VI, in this example, there are two buildings BLDa and legs at the camera DS =. In the photo pic, the two buildings are photographed at the angle of v: degrees, but the heights of the buildings are different, as shown by the side = V2, ′ are Ha and Hb, respectively. The photo in the side view indicates the photographed area of the building with a ^ line. The top of the thick line is of course the line connecting the DSC of the camera and the top of the building. , ..., Ben? Ming Department recreated the positional relationship shown in Figure 2 in a virtual space. At the beginning of modeling in the virtual space, the object BLDa and its subsequent height are unknown. But a, if the camera DSC and photo PIC are configured according to the position relationship with the photographic time, there should be makeup of buildings BLDa and BLDb in the extension of @ ί 物 (thick line in the picture) displayed by PIC: 5 In the virtual space, the top of the pattern which is parallel to the moving plane shape in the height direction overlaps with the top of the photo, and the position relationship shown in Fig. 2 can be displayed in the virtual space. Therefore, without measuring the actual height of the building &, BLDb, the respective heights can be specified in the virtual space. The above shows the principle of three-dimensional modeling in the present invention. Here, in order to make the principle easy to understand, it will be described according to the specific examples shown in FIG. 丨 and FIG. 2, but the present invention is not limited to these contents. In the following, the concept of the present invention constructed according to this principle will be described. The second 7C modeling method of the present invention is composed of the steps shown below. Page 8 2061-4944-PF (N); ahddub.p t d 565810 V. Description of the invention (5) f Step (a) of the present invention, enter the phase of the building represented by electronic data =. As for the input of photos, for example, electronic data obtained by using a digital still camera can be input, or silver or lead photos can be input by using a scanner or the like. This photo does not necessarily cover the entire building, at least the top of it is sufficient. In step (b), input the photographic position of the photo and the phase of the building = standing relationship and the actual second in order to make the building look from the photographed position: The photo is suitable for the required photographic parameters. The relative positional relationship and the shirt parameters are used as data to reproduce the state of the shooting in the virtual space. Because: It is not necessary to define the photographing position by absolute coordinates such as latitude and longitude, but only according to the relative position relationship of the built object. Of course, absolute coordinates such as latitude and longitude may be used for photography and building. The shooting parameters can use various parameters that can specify the direction the camera faces when shooting and the image angle of the camera. Fig. 3 shows the image angle of the camera where ΐ = 1 and the distance between the lens L and the CCD of the photographing unit is the focal length f. According to the focal length, the artifact, that is, the width of the range 0BJ that can be photographed by the CCD of the photographing section. Therefore, the image angle 0, the focal length f, and the like can be used in the parameters of the special image / image angle. FIG. 4 is an explanatory circle showing the shooting direction of the camera. The Y Z axis in the figure corresponds to the axis in the virtual space. In micro-viewing, because the latitude Ϊ confronts Ξί, it can be orthogonal coordinates, so that the x-axis, γ-axis and latitude, and the line's second degree are defined in the 3rd axis of the virtual space, and the declination angle can be based on this: : ..., roll angle, specific camera's shooting direction Kang Bi depends on a variety of parameters to determine the shooting direction, but if you use the set shown in Figure 4; legislation, description of the invention (6) degree 'Am parameter value is also easier to obtain After the number of materials and materials, in step (0, the three-dimensional plane shape and photo are generated: the position specified by the building is defined according to the relative position relationship. Jian = & set, and the photos are not necessarily completely specified according to the shooting parameters The surface shape is shown in the examples of Fig. 1 and Fig. 2. The plane shape of the corresponding part may be y, and the line of the photographing face may be specified. W. Therefore, the plane shape is a line segment or folded in a step ( d), 蕻 基 产 一 & photos-until the specific g should be built ▲ To map the plane shape to the operation of the operator 彳 ^: Do not shape in the direction of the south. The mapping is based on this It can also be done automatically. It can be highly efficient: ^ Typed squares & You don't need to measure each building much higher than just modeling by the building and the second time. Moreover, you can achieve three-dimensional modeling in degrees. 卩 White number speculation 咼The essence of virtual methods

In the case of multiple buildings in accordance with the second aspect of the present invention, the second method: the modeling method, which enables three broadcasts to be carried out at the same time in two photos, including buildings, and can be performed for multiple buildings according to the one photo. Advantages. Photography is not necessarily three-dimensionally modeled on the ground. Two pieces can be easily implemented in a wide range as described above. The plane shape used is not necessarily: the three-dimensional modelling method of the present invention makes the two-dimensional plane map specific = two specific, but the basis includes buildings. Achieve high accuracy: A-乂. By using a highly accurate flat shape, ~ 7L modeling at a time. 2061 -4944-PF (N); ahddub.p t d 565810 V. Description of the invention (7) In the three-dimensional modeling method of the present invention, two steps are defined to represent the rough shape of the building, and the two models are included. Buildings are not only more complex shapes such as the secondary-like houses of a large building. The basic three times =: there are also f according to the building,-general houses are classified after the buildings are defined: 糸 like this model. In the definition of a planar shape, the shape includes a rough shape in the direction of two degrees, which can also be modeled in two dimensions. It is also possible to perform the third-class freedom with accuracy and less labor, but use the electric analysis to determine the edge of the building; 2: analyze the edge of the plane shape and the edge of the plane according to the analysis of the photo: process The overlapped state becomes a maximum map: the steps of ί; =; and by dividing such noise data, the specific reality is determined; as long as the negative map becomes the maximum map. The near test? The second one: the modelling method ’hopes to reflect the low of one of the buildings to the south. By inputting the height of the surface near the building = low data, the high and low data are reflected on the surface of the building near the building to take into account the shape of the high and low data. For example, there is a method in which the shape of a specific building in the upper part of the surface above and below the surface where the high and low data is reflected. By doing so, the height of a specific building can be more accurately. In the three-dimensional modeling method of the present invention, in addition, in step (6),

2061-4944-PF (N); ahddub.ptd page 11 565810 V. Description of the invention (8) At least one of the specified ones is prepared to be pasted as a knot and processed in a field similar to this. Example The ease of repeating regions and bit structures. In addition, the surface itself can be structured in front of the object, which can be integrated in the entire area. However, the present invention can also be used as a high-level square of the dimension electrons as a comparison. The photo repeating area of trees and the like can be compared, or the photo structure can be pasted to produce the shape of the map data. The photo with a certain height t photo ’3 ° $ On how to’ It is easy to improve the authenticity of the model by the surface. Patches: The case of the structure ’I want to divide into repeating areas that are repetitive and non-repeating areas =: & I repeat the domains and separate areas, and paste on the sheet whether or not to obtain the actual structure-independent repeats. If so, the paste of the repeating area becomes: the situation. As shown above, * repeated objects = part of the photos obtained should be used; Fu Gu Yi avoids the impact of obstacles. The separate three-dimensional modelling of the hard building consists of the method of generating the three-dimensional electronic ground of the building. Second, the present invention can also be constructed as a three-dimensional modeled support device for supporting a three-dimensional model of a building. For example, the modeled support device for this material may include an image data input unit, a horse-dimension dimension unit, a virtual space display unit, a projection unit, a model display shirt input unit, and a photographic information input unit. Parameters of flat image data. The virtual space display unit displays a virtual < shirt image in the direction of the building viewed from the photographing position in the three-dimensional electronic photography space. projection

565810 V. Description of the invention (9) According to the part 'display in the virtual m, it can be carried out efficiently by mapping the plane shape in the height direction 3 ::: structuring, support of the two-dimensional model modeling Modeling. Automatically generated; ί: In addition to the structure of the aid device, the present invention also includes the formation of secondary electronic data of biological objects. In this case, for example, the model input section, the photography information Qiu device explained. Modeling Department ^. In addition, it also includes the definition of the modelling department's chess player in the virtual space Θ, the deployment of photos ># 七 i 十 面 形 and photos-until, ^; in the direction of the game The flat shape reflects Qi Nai Shuang Ma Zhi, the specific building in the photography side of the invention can also be used as a collection of data for the direction of the shape of the data collection equipment: Second: Lei Zao caide Department and data misstore department. Take a photo of the building with the electronic shellfish. I 丨 Camera shirt The shooting parameters required to make the built-in film 3f look from the shooting position of the photo. Data storage department: The electronic data and photographic parameters are stored in relation to the data; Off: ΐ material collection device, which can be collected in a unified way, but I hope that the digital parameter acquisition department is built in: Knife. Especially posture and focal length. The device is used for realizing the above-mentioned display structure by using a computer when taking pictures for taking pictures. The electronic assets can be included in the form of acquisition and the position of the phase shadow. It is hoped that the camera will be modeled in terms of dimensions 565810 Fifth, the description of the invention (10)? Computer program configuration of supported makeup is also possible. In addition, the recording medium structure of some programs may be used. In addition, in the recording medium ^, you can use floppy disks or CD-ROMs, magneto-optical discs, IC cards, _cards, η cards, printing & symbols printed with bar codes, and internal memory devices (ram or _ Etc.) and various media readable by computers such as external storage devices. Embodiments of the Invention The embodiments of the present invention will be described in terms of the following items. Α · The structure of the map data generating device: A1 · The structure of the automatic modeling section: A2 · The structure of the manual modeling section: A3 · The structure of the structure generating section: B · The data collection device for modeling: C.Modeling: C1 · Pre-processing: C2 · Automatic modeling processing: C3 · Manual modeling processing: C4 · Structure generation processing: C5 · Additional structure setting processing: A · Structure of map data generating device: Figure 5 shows the third embodiment Meta Electronic Map Data Generation System

2061 -4944-PF (N); ahddub.p t d page 14 565810 V. Explanation of the structure of the invention (π) 1 0 0 (hereinafter sometimes referred to as system 1 0 0). After the three-dimensional modeling of the building in this system 100, three-dimensional electronic map data is generated. In this embodiment, the system block 100 is constituted by a function block that constructs an icon in software in a computer. In the following example, a case where a system is configured using a separate computer is described as an example, but a structure in which a host computer and a terminal are connected via a network may be used. Each function block in the system 100 'has the following functions. The data input section 101 imports the modeling data from the outside. The data for modeling is the data required for the three-dimensional modeling, including the electronic data of the photos of the building to be modeled, the photographing position, and the photographing parameters. The photographing parameters in this embodiment use a tilt angle, a pitch angle, a roll angle (refer to FIG. 4) indicating a camera direction during photography, and a focal length (refer to FIG. 3) indicating an image angle. These data can also be input through media such as optical disk M0, and can also be input via communication such as the Internet. According to the input data for modeling, each function block shown below performs three-dimensional modeling. The pre-processing unit 110 has a function of preparing a virtual space used in the three-dimensional modeling in the system 100. More specifically, it has the function of preparing the surface of the ground before the quadratic model is generated. Pre-processing section 丨] 〇 To perform this function, refer to the elevation database 103 and the two-dimensional map database 102. The elevation and material library 103 is a database for assigning marks to points on the ground surface that are cut into nets and meshes. If it is in Japan, for example, you can use the 50m mesh numerical data of the National Institute of Geography. The two-dimensional map database 102 is an existing map database on the plane. Therefore, the flat shape of each building can be considered. On the two-dimensional map database 102, for example, aerial photos,

565810 V. Description of the invention (12) Star photos, residential maps and other materials. The automatic modeling section 120 has a function of automatically generating a three-dimensional model according to the modeling data. The virtual space display unit 104 is used to display the modeled virtual space on the display of the system. The principle of automatic modeling is shown in the description of Figures 1 and 2. In the automatic modeling section 丨 2 〇 As will be described later, a function block prepared for automatic modeling using this principle is prepared. The manual modeling section 130 has a function of generating a three-dimensional model in accordance with an operator's operation. In this embodiment, a complicatedly shaped building or the like that cannot be fully automatically modeled becomes a modeled object. The virtual space display unit 104 provides display of the virtual space that the operator should refer to when manually modeling. As will be described later, the manual modeling section 130 prepares functional blocks to support manual modeling. The structure generating unit 140 has a function of attaching a structure to the modeled building. In this embodiment, a picture of a building is cut out as a structure. The structure is pasted by referring to the display on the virtual space display section 04. The structure generating section 140 prepares function blocks for pasting to support the structure. The additional structure setting unit 150 models the building required for the electronic map data other than the building. Additional structures include, for example, trees, signals, guard rails, and the like. Use the model registered in the parts library to define additional structures. The integration unit 106 has the function of adjusting the data generated in each of the above functional blocks into a format of electronic map data. Also set the text setting of the building or place name, and the setting of various signs to be displayed on the map. The integration unit 106 outputs the integration to the map database 10 as such

565810

ME's recording media ME. The electronic map data can also be recorded in DVD-ROM and other types A1. Structure of the automatic modeling section: Figure 6 shows the structural modeling section 1 of the automatic modeling section 丨 2 0 as shown above, according to Figure 1 And the function block shown in Figure 2 ° automatically performs the three-dimensional modelling of the building. J = The gray level difference of each side of the building is extracted from the photograph of the building by the drawing J 121. The extraction of the margins can be performed, for example, by selecting the basic model to be used from the basic model database 123 in accordance with the basic model selection section 122. In the figure-and the basic model is shown as an example. ^ The model includes the model MD1 of the building, the model MD2 of the bungalow, the MD2 of the second floor / Chu type MD3, etc. The base model is a rough shape 赋 that gives a building like this. In this embodiment, the basic model is selected in a fixed order until a proper result is obtained in the comparison processing described below. It is also possible to select the basic model according to the information provided by the 22-2 map database. … ~ ^ 70 ^ The comparison unit 1 24 performs the comparison process, that is, the process of changing the ancient model of the base model to make it suitable for the photos arranged in the virtual space. Judge whether it is suitable according to the overlapping state of shadow silly ^ fate and basic model. In the case where the height of the base model is not suitable either, select the base model to communicate the message to other base models, and then compare the other base models. The ° line skeleton generating section 1 2 5 generates a code for a building according to the result of the comparison. ’, ,,, month

565810 V. Description of the invention (14) A 2. The structure of the manual modelling unit. Figure 7 shows the structure of the manual modelling unit 13. The unit UG uses the function blocks shown below to generate manual operations. Three-dimensional model. f Stand-up shape selection section 1 3 2 Select a stand-up shape in accordance with the operator's instructions. The standing shape database 133 is defined in advance. An example of the standing shape is shown in Fig. ^. The standing shape means a model shape defined by moving a planar shape in a virtual space to a height direction. For example, it can include single, vertical ascent Rpl, point concentrated ascent Rp2, line concentrated ascent Rp3, zoom in ascent Rp4, stage ascent Rp5, dome ascent Rp6, and so on. By defining various standing shapes in this way, it is possible to cope with the shapes of various buildings. The height changing section 1 31 moves in the height direction of the planar shape in accordance with the selected rising shape. In order to perform more complicated modeling, the stack processing unit 134 processes the overlap between the formed models and the like. The wire skeleton generating part 1 3 5 generates the wire skeleton of the building according to the processing results of the height changing part 1 31 and the stack processing part 1 3 4. A3 · Structure of the structure generating unit: Fig. 8 is an explanatory diagram showing the structure of the structure generating unit 140. The structure generation unit 140 uses the function blocks shown below to apply a structure to the modeled building. The repeating structure setting unit 141 defines a repeating area in which the same structure is repeatedly used for the building and other separate areas in accordance with the operation of the operator. 2061-4944-PF (N); ahddub.ptd page 18 565810 V. Description of the invention (15) In addition, the line skeleton of the building is cut according to the setting of the repeating area. The structure setting unit 142 sets a structure to be attached to each area. In this embodiment, the structure is set using a part of the photo. Structural correction department! 4 3 Modify the structure obtained from the photo cut into a surface suitable for the model. A correction example is shown in the figure. The hatching in the figure indicates the structure obtained from the photograph. The correction method includes the correction of the shape of the structure, and the combination and separation of multiple structures. It is also possible to adjust the color including the structure. The structure affixed to the upper part 44 is affixed to the structure set according to the definition of the repeated structure. For repeated areas, the structure set by repeated application is applied in separate areas. The structure paste upper part 1 4 4 stores the pasted result as a structure database 145. The structure database 145 is a database corresponding to a code defining a structure and a code defining a face pasted on the structure. B. Modeling data collection device: Fig. 9 is an explanatory diagram showing the structure of a modelling data collection device. The data collection device for modeling is composed of a digital still camera (Dsc) and data; ㊁2.7. For the data recording device 27, for example, a GPS 25 can be mounted on the DSC20 body using a portable force. GPS25 is like a device that uses radio waves from an artificial satellite to set # 要 夕 ^ as well-known as a private device that only matches the position. Synchronized acquisition of this shutter & shutter 21 Li Xuyi Λ ★ + Beijing Example, quarter% pay, year, longitude, height information The body of the fDSC20 is also provided with an angle sensor 24. Angle production "Bu 24 and shutter 21 are acquired synchronously. Sensing the state of motion

565810 V. Description of the invention (16) ^ Pitch angle, roll angle, the angle is obtained by the rotation angles around the X axis, Y axis, and Z axis in the figure. The under mount is a focal length sensor 23 provided on the lens 22 of the DSC 20 and the focal length of the shutter 21. Ding Ban ~ Yin Zhi 铮 each; = The obtained data is transmitted to the data record 7 through the interface 26 and stored in the shell material recording device 27 as a photographic database. In the figure-and the structure of the photographic database 28 is shown as an example. In the picture, the strontium die Z? Each photo is assigned the code number "ZN00 1 00" as the key, = latitude "LAT1", longitude "coffee", focal length "⑽" photo, 71 ". It is also possible to save the photo taken with it = shell material, other annotations, etc. In this embodiment, the case where each sensor is incorporated in the DSC20 is shown as an example. It is also possible to include the sensors separately from the DSC20. On the separate included structures, for example, the structure of each sensor mounted on a fixed DSC20 tripod, etc. · 3D modeling: Figure 10 is a process diagram showing the 3D modeling. Here, the processing contents in the case of modeling a building are shown. The system 100 shown in Fig. 5 performs each process automatically or in accordance with the operator's operation. Secondly, in the two-dimensional modeling, first, the operator inputs modeling data in the system 100 (step s 100). The modeling data includes the photographic parameters and photos shown in FIG. 9. Second, the operator models the building based on the data. According to the virtual

565810 V. Description of the invention (17) Pre-spatial processing unit (step S20 0), automatic modeling (step S300), manual modeling (step S40 0), structure generation (step S500), additional structures The order of setting (step S60 0) is modeled. After the system 100 registers the model generated in this way as map data (step S700), the three-dimensional modeling is completed. Each process will be described below. C1 Pre-processing: Fig. 11 is a flowchart showing pre-processing of a virtual space. This process is equivalent to the function of the pre-processing unit 110 (see FIG. 5). When the process is started according to the operator's command input, the system 1000 inputs the shape, position, and elevation data of the building near the model (step S 2 0 1). The second-dimensional map information of the building is used for the shape and location of the building. However, aerial photos can also be used. Records at the photography site can also be used. In this embodiment, the planar shape of the building is input as a completely closed graphic based on the two-dimensional map data, but only the part appearing on the photo may be output. For example, when the photo is not taken on the back side, these parts may be omitted, and a polygonal input plane shape may be used. In this case, by individually generating the front face model and the back face model, the two can be combined to complete the building: dimension model. The position of the building may be specified by an absolute coordinate system such as latitude and longitude, or may be specified by a relative relationship with the photographed position. In the present embodiment, the data of the objects near the building and the object are used in accordance with the data of the South China Institute of Land Use, but it is not limited to this. Secondly, the system 100 defines a virtual space as an infinite space (step S202). At the same time, as shown in the figure, I want to show the

2061-4944.PF (N); ahddub.ptd Page 21 565810

V. Description of the invention (18) The situation of infinite space. Three-dimensional graphic display in virtual space can be applied with well-known techniques. The position of the viewpoint can be arbitrarily set during the three-dimensional graphic display, but it is convenient for subsequent processing to make it consistent with the shooting position. When the infinite space is set in this way, the system 1 00 configures the two-dimensional shape of the building according to the data input in step S20 (step S203). In the figure, the box with a thick line indicates the two-dimensional shape of the building. Finally, the system 100 adds undulations to the infinite plane according to the elevation data (step S204). At the same time, the pre-processing is completed by making the secondary shape of the building suitable for the undulating ground. FIG. 12 is an explanatory diagram showing the meaning of the additional fluctuation. Shows the state of the building and camera viewed from the side. As shown in the figure, the ground surface has a fixed elevation near the building and near the shooting position, that is, the ground surface is not consistent with the horizontal plane. Let the camera DSC elevation be ALTc, and the elevation of the building be ALTb. ,Do not. It is known from the principle shown in Figs. 1 and 2 that the modeling method in this embodiment is not the height of the ground from the building, but the position of the top Tb of the building. This position is determined relative to the DSC's photographic position. Therefore, the undulations of the ground surface are not considered in the virtual space, and by setting the camera DSC to the elevation ALTc, the position of the top Tb of the building, that is, the elevation can be set with high accuracy. However, in this case, the model of the building is also included between the actual ground surface and the horizontal plane (the hatched area in the figure). Due to the difference between the height of the building and the actual situation, this model cannot be said to reflect the reality with high accuracy.纟 Example, by adding undulations to the ground surface in advance, and disposing the building on its surface

2061-4944-PF (N); ahddub.ptd page 22 565810 V. Description of the invention (19) The quadratic shape of the object, to avoid this disadvantage. In this embodiment, after the undulation is added to the ground surface, the building is modeled, but on the contrary, after the building is modeled, the elevation may be considered and the oblique line part in the figure may be deleted from the model. If the difference in elevation between the photographed location and the building is relatively small, it may be possible to omit the principle of adding undulations. C2 · Automatic model processing ·· 2 in the model module = 〇〇 ^ t Set in the virtual space // without photo configuration. The origin θ is the image angle θ at the time and: a 1 ί camera position. The shooting is determined according to the shooting parameters, the image angle of Zhao is 0 and the shooting direction. ’,,, yaw angle and pitch angle of the point chain line table does not follow the vertical line of the photography line and the direction of the ambiguous photography shirt at both ends of the photo. The picture is based on the -dot chain (㈣, Ε2) as: ^^^ The partial roll angle when making the movie is placed around the cow and placed in the virtual space. The photos are shown without tilting in the photo. '"胄 Λ Turning' makes the building in the virtual space Xinyi S304). In the figure-with the collocation / when the system 1GG extracts the edge of the image (as shown on the left side of the step, the building does not extract the edge of the image. As shown in the figure, the boundary is not gray. The drawing of the edges = 3 lines on each side with the edge part as the boundary method. There is no limit to 14 types of processing, and various 2061-4944-PF (N) can be applied; ahddub.ptd Page 23 565810 V. Description of the invention (20) After the extraction of the edges is completed, the system 100 compares the basic model, that is, the determination of the overlapped state (step S30 6). The basic model is selected / shown in Figure 6 Rough shape. In this embodiment, try to compare after selecting other parts. In the case where the proper comparison result cannot be obtained, the J model is not selected properly, and the next foundation is selected according to the preset order: soil. Fig. 14 is an explanatory diagram showing a method for judging an overlapping state. The steps are shown on the side. The left side of the figure shows a model with an edge from the object's building. The system 100 is shown in the right side of the figure. Ancient Cheng PI, P2 ... The overlapped state of the extracted edge and / or the segment contained in the shape. For example, regarding the height p3, the overlapped state of the part indicated by the arrow. Β τ The determination of the overlapped state can be performed by the method shown above the figure. The area 2 A1 is defined with a fixed width around the edge L1 indicated by the solid line. Similarly, as shown by the dotted line, the area A2 is defined with a fixed width around the edge L2 of the base model side. According to the position of the edge li and the edge L2 In areas A1 and A2, overlap occurs in a consistent range (the oblique line in the figure). The area of the overlap changes with the height of the base model. For example, when the height of the base model becomes higher, as shown from the left in the figure, The ^ product of overlapping parts was enlarged once and then decreased. This change is graphically represented in the center of the figure. The consistency rate is the area of the overlapping part relative to the sum of areas A 丨 and A 2. As shown in the figure, the basic model is used. When the height becomes higher, the coincidence rate gradually increases =, the degree P6 becomes a maximum value, and then it decreases. The system judges the base model and the edge at the height p6 according to the change of the coincidence rate like this. 1-4944-PF (N); ahddub.ptd Page 24 565810 V. Description of the invention (21) Of course, the judgment of 'comparison after subdividing near the height p6 can also overlap. The judgment of sadness is not limited to In this example, various methods can be applied. I. When the comparison is completed, the system 100 specifies the height of the base model according to the result, and generates the line skeleton of the building to be the object, S308). C3 · Manual model processing: FIG. 15 is a flowchart showing a manual modeling process. This is equivalent to the function of the manual modeling unit 13 (refer to FIGS. 5 and 7). Step s402). The photo placement process is the same as the automatic modelling process (step 2 in Figure 2). The display in the virtual space is less than the display of the system 100. 々 乃 行 立 2 = According to the plane shape and relative appearance in the virtual space, See the author to choose i ratio Γ / 7_). Regarding the standing shape, after selecting the height in the virtual space, the system will generate a model of the height and display it in a variety of ways. However, the display of the space can be used to operate the pointing device such as a mouse. # 望 For example, automatic manual modeling in virtual manual modeling. Buildings with complex shapes are also available; U-mode & early-purity comparison cannot be processed Page 25 2061-4944-PF (N); ahddub.ptd 565810

= Case. In the manual modeling, in order to form such a complex I-± gold β ^, a stacking process is performed (step S406). In the figure: Summary of table processing. For example, as shown in the figure, the stacking process is used to advance the shape of the material in the upper and lower parts. This kind of building completes a model by combining the model of the lower part and the surname of the upper part. Stacked processing means processing of multiple models like this. The combination of models is not necessarily limited to overlapping squares, but the combination of models adjacent in the horizontal direction is also acceptable. FIG. 16 is an explanatory diagram showing a method of stack processing. Shows the daytime display of the HDISP display of System 100. The three dimensions of the virtual space are displayed on the left. 4 shows V L and V R is displayed on the right display plane. Pictures of buildings are omitted. Regarding the two-story structure shown in FIG. 15, the upper-floor part ^ planar shape OBJin and the lower-floor part planar shape OBJut are preliminarily given. The operator moves the planar shape 0BJout of the lower part in the height direction, and as shown on the left, a model LE1 of the lower part can be generated. In this embodiment, at this stage, the planar shape OBJ i n of the upper layer portion is set so as not to affect the planar shape OBJout of the lower layer portion. That is, it does not matter whether or not the plane shape OBJin is set. Secondly, the operator can generate the model LE2 of the upper layer by moving the planar shape OBJin of the upper layer in the height direction. In this embodiment, the model LE2 of the upper layer is formed from the model LE1 of the lower layer. By such a stepwise modeling, modelling of a multi-layered building can be realized. The stacking process in this embodiment is just an example. For example, the model U 2 in the upper part is set to a hollow state, and the mold SLE 1 in the lower part is formed from the ground surface.

565810

After the model LE2 of the upper layer part is formed on the surface, the two are combined. In addition, the model LE2 of the lower layer part is generated as the same as the embodiment. After the model LE1 of the upper layer is formed from the ground surface, it may be combined with a B-type logic j shape. In the example shown in Fig. 16, the lower part and the plane shape are known in advance. In the case where the planar shape of the upper part is ωBJin, which is unknown, the three-dimensional display of VL or planar display of VR makes the operator's emotional shape acceptable. This definition is also performed before the modeling of the lower part. It is also performed after the modeling of the lower part and before the modeling of the upper part.

When compared and stacked in this way, the system is 100 °. After the height of the base model is specified, a target result is generated (step S408). Card early line skeleton C4 · Structure generation process: Figure 17 is a flowchart showing the structure generation process. This corresponds to the function of the button 140 (see Figs. 5 and 8). In the structure generation process, the operator first sets the repeating structure (steps, that is, the repeating area of the same structure and its separate areas can be repeated. An example of setting the repeating structure is shown in the figure. The part of the additional line is equivalent to the repeating area. The structure of each step in the building and the like is similar. The repeating area can be set in units of each step. The part other than the repeating area is defined as a separate area. Because the repeating area in this process is defined In order to use the structural unit, the heavy product domain may not correspond to the actual structure of the building. For example, as shown in the figure

565810 V. In the middle of the description of the invention (24), there is also a multi-level repeating structure on the front, but the side is composed of one side. In such a building, it is also possible to set a repeating area in units of steps for the front side and a separate area for the side face. When the definition of the repeat structure is completed in this way, the operator sets the structure (step S504). Generate a structure from a photo of the building. It is also possible to use a photo that is the same as the one used in the three-dimensional modelling. But in this implementation ^, considering the burden of fitting the structure to the model, it was decided to use the photo from the front of the building. In the structure setting process, the operator cuts out a part of the photo into a structure, and corrects the shape and hue to make it fit the surface of the model. This amendment also includes the combination and separation between structures. When the setting of the eight-birth photo frame is completed, the system 100 pastes it according to the operator's order (step S5 ° 6). The model for the repeating region is shown in the figure. The right side of the photo ’is the model created. The top-level photos can be pasted on the building to complete the whole structure of the building. If this = trees, obstacles, it is often rare to use the highest level of the building to take pictures. ^ ^ ^ Define the repeating area, which can be beneficial to the structure of the building as a whole: Department: the structure '▼ It is easier to construct, However, the areas that are processed separately are also pasted with repeated knots. The structure can also be affixed to a separately prepared structure. C 5 · Additional structure setting processing: The acoustic system function of the additional structure setting place 150 (refer to FIG. 5) is modeled in the additional structure setting section & management ’system and the building itself.

565810 V. Description of the invention (25) Same as the surroundings of the righteous building. Use the operator's operation: the side treatment. In the present embodiment, the additional structure is based on the following: a model of an additional structure provided with equipment. Pieces. Urine, and layout designation methods are classified into point parts and line zero designation examples. A point part is a part that can be modeled by including a point on it. In the small guest, change to the machine, street lights, telephone poles, post boxes, bus stations, railways, mobile phones, etc. Include the shape of the high-voltage electric wire system on the large point parts. , * T towers, gas stations, convenience store chains, etc. < y 24 Also point parts may not necessarily be available according to the modeling method shown in this embodiment 'available: general three-dimensional modeling method. Figure 2 shows the configuration of the traffic signal on the instance of the point type part. As shown above, when the display DISP 'displays the setting of the virtual part on the left, select the magic "ΪΓ ί 千 面Display the VR. Select the type of ▼ pieces at point zero, Jinwei-M, and specify the set point of the part with νΐ: in the flat display VR. Then, display the VL display model in three dimensions. Addl. The height of the tree can be set to VL. You can set the height of the tree. 隹 一 -Man Wuxian ... Description of the setting example ®. The line part means that the kind of part includes the m-retaining wall and the sidewalk white line tower. ^ Of = white line , Various bridge shapes, central isolated islands, etc. These models can also be formed by the second 70 modelling method. The line unit registration unit length is divided into ^ bit models, and the unit model is used repeatedly according to the specified range | _ page 29 2061-4944-PF (N); ahddub.ptd 565810 V. Description of the invention (26) In the figure, the day-to-day display of the condition of the maintenance license is shown on this line part instance. VR is displayed in one element, such as The arrow indicates that by the specific line segment range, that is, the starting point of the line segment At the end point, yL is displayed in the three-dimensional model as Add2. D. Effect: According to the electronic map data generation system of this embodiment described above, the three-dimensional model can be performed without measuring the orientation of the table. Therefore, the building can be modeled with relatively little labor and high accuracy. The processing after Figure 10 illustrates the modeling of a single building ', but multiple buildings are photographed in one photo. In this case, these buildings can be modeled in parallel. Therefore, three-dimensional electronic map data can be generated more efficiently. First, in this embodiment, by using the structure according to the photos of the buildings, the authenticity of the model can be further improved. X. By using repetitive structures, ^ applies these structures with a lighter burden. E. Modifications: In this embodiment, 'example' shows the use of a model of photography on the ground. If the shooting parameters and shooting positions are already It is known that the 4-mesh film is not limited to photographing on the ground. For example, the self-climbing direction can also be used: " membrane. By using known photography parameters and photography position == aviation can be carried out efficiently Three-dimensional modeling. In this embodiment ', the modeling is performed through the steps shown in FIG. 10, but

2061-4944-PF (N); ahddub.ptd page 30 565810 V. Description of the invention (27) It is only necessary to omit the automatic or manual modeling. In addition, the structure generation process and the additional structure setting process are processes performed to improve the authenticity of the three-dimensional model, and it may be omitted according to the purpose of the model. Although various embodiments of the present invention have been described above, the present invention is not limited to these embodiments, and various structural systems can be employed without departing from the scope of the present invention. For example, the above control processing may be implemented in hardware in addition to software.

2061 -4944-PF (N); ahddub.p t d page 31 565810 Simple illustration of the diagram Figure 1 is an explanatory diagram showing the principle of three-dimensional modeling. FIG. 2 is an explanatory diagram showing a relationship between a photographed image and a building. FIG. 3 is an explanatory diagram showing an image angle of a camera. Fig. 4 is an explanatory diagram showing a photographing direction of a camera. FIG. 5 is an explanatory diagram showing the structure of a three-dimensional electronic map data generating system 100 of the embodiment. FIG. 6 is an explanatory diagram showing the structure of the automatic modeling section 120. FIG. 7 is an explanatory diagram showing the structure of the manual modeling section 130. Fig. 8 is an explanatory diagram showing the structure of the structure generating section 140. FIG. 9 is an explanatory diagram showing a structure of a data collection device for modeling. FIG. 10 is a process diagram showing three-dimensional modeling. FIG. 11 is a flowchart showing processing before a virtual space. FIG. 12 is an explanatory diagram showing the meaning of the additional fluctuation. Fig. 13 is a flowchart showing an automatic modeling process. Fig. 14 is an explanatory diagram showing a method for judging an overlapping state. Fig. 15 is a flowchart showing a manual modeling process. FIG. 16 is an explanatory diagram showing a method of stack processing. Fig. 17 is a flowchart showing a structure generation process. Fig. 18 is an explanatory diagram showing an example of setting a point part. Fig. 19 is an explanatory diagram showing an example of setting of a wire part. Explanation of symbols 1 0 ~ map database; 2 shutters; 2 2 ~ lenses; 2 3 ~ focal length sensors;

2〇61-4944-PF (N); ahddub.ptd page 32 565810 Schematic description of 24 ~ angle sensor; 2 7 ~ data recording device; 100 ~ electronic map data production 1 01 ~ data input section; 103 ~ Elevation database; I 0 5 ~ Parts database; II 0 ~ Preprocessing section; 1 2 1 ~ Image edge extraction section 123 ~ Basic model database 1 2 5 ~ Linear skeleton generation section; 1 3 1 ~ Change section 133 ~ shape database; 1 3 5 ~ wire skeleton generation section; 1 41 ~ structure setting section; 143 ~ structure correction section; / 145 ~ structure database; 2 6 ~ interface; 28 ~ photography database; production system; 102 ~ 2D map database; 104 ~ Virtual space display unit; 106 ~ Unification unit; 120 ~ Automatic modeling unit; 12 ~ 2 Basic model selection unit; 124 ~ Comparison 1 3 0 ~ manual modelling unit; 132 ~ shape selection unit; 1 3 4 ~ processing unit; 140 ~ structure generation unit; 1 2 2 ~ structure setting unit; 144 ~ structure attachment part; 150 ~ additional structure物 定 部。 Object setting section.

2〇61-4944-PF (N); ahddub.ptd p. 33

Claims (1)

565810 VI. Dimensional modeling method of patent application data, generating three-dimensional electron (a) of the building, entering the step relationship of the photo of the building represented by the electronic data, and the photographic position of the photo and the building Relative location photos, Shang Ershang: The photographic location is based on the actual landscape of the building and the steps that are suitable for the required photographic parameters; the phase oblique 0) in the virtual space where the three-dimensional electronic data is generated, according to The 2 position relationship defines the planar shape of the building and the photographic position ': and = the photo is arranged in accordance with the photographic parameters so that (ί) / Λ in / degree direction J1㈣ planar shape is mapped to the photo First, the steps of the shape of the temple ^ building in the height direction. 2 For example, the three-dimensional modelling method of the patent application scope, the broader plane shape is based on the two-dimensional plane map including the building. 3 ^ Three-dimensional modelling method of the scope of application for patent β, its outline: Shape :) US package /: This plane shape together defines the three-dimensional element that represents the basis of the rough shape of the building Steps of the model. = In the three-dimensional model of the μ scope of the patent application, step (d) includes: eight steps of analyzing 4 edges of the building based on the photo t; steps of overlapping states; and analyzing the The edge of the plane shape and the ridge line become the maximal step of mapping the specific height by the overlapping state. Page 34 2061-4944-PF (N); ahddub.ptd 565810 6. In the scope of patent application, 5 is in the patent application The three-dimensional modelling method of the first item in the scope, which previously included rotating data indicating the height of the land near the building; ππ ^^ (c) also includes orders to the surface near the building Reflect the steps of the ancient low shell; ；, the shape of the side. The step (d) specifies the shape in the height direction than the upper part of the surface, the three-dimensional modeling method of the first scope of the patent, complex: 今 ίίί This step (d) is used to specify the step of attaching the phase shape on the surface of the model of the table with makeup in the height direction. A < Master less. Divided as the structure 7. If the patent application scope No. 6 In item two, this step ⑷ Including:-the type method, which defines the area on the surface of the structure and a separate area outside the repeating area :: single: the focus of the structure is on the repeating area and where the photo is ... The steps of the structure of the unit structure. The structure that is not related to the structure of the unit 8—a kind of three-dimensional electronic map data of the three-dimensional electronic map data, including the method of generating (a) the input of the electronic data. (b) Enter the steps to specify the photos of the two objects on the map; edit the photographic position information of the photos and the steps to specify the photographic parameters of the photographic photography <photography; Camera direction and image angle (c) According to the data obtained in steps (a) and (b), 2061-4944-PF (N); ahddub.ptd p. 35. The scope of patent application is obtained from the flat map, and the two are placed in the height direction: and the step W of the photo ... The shape of the building in the high-floor area is mapped to the shape of the three-dimensional model rt I connected to the photo. The production of zibei materials includes: • The device 'supports the three-dimensional image data input of the building * photo and plane shape; input the phase C of the building's phase represented by electronic data, and enters the phase ... The photographic location of the film and the building, the scene and the photo are suitable for the location: the location of the movie to see the real virtual space of the building, display the neighbourhood, the photographic parameters; in the room, the virtual three-dimensional electronic data is generated from the photographic location. The plane shape if large is large, and it is obvious that the second sister configures the image according to the relative positional relationship; and ',,' does not look at the projection direction of the building from the shooting position, and the parameters of the burial foundation are ^ = The photo is arranged in the virtual space according to the location; and the position 'projects the photo to the displayed image model display section + the 9 of the building in the ancient meaning, and then maps to the plane shape. 10—A The status is displayed in the displayed image together. Sub-data. Including two, meta-modeling device ’to generate a three-dimensional electrophotograph of a building: Department’ enter the electronic representation of the building Page 36 2061 -4944-PF (N); ahddub.ptd 565810 Patent application scope photography information input unit, input the relative positional relationship between the photographic position of the photo and the building and in order to see the building from the photographed position The actual landscape and the photograph are suitable for the required photographic parameters; and the virtual space display section defines the planar shape and the photographic position in the virtual space in which the three-dimensional electronic data is generated, and the photograph is taken. It is arranged at a position specified by the photographing parameter, and maps the plane shape to the shape of a specific building in the height direction by the direction of fn degrees according to the photograph. Medium 1 1 · If the three-dimensional modeling device of the scope of the patent application, the modelling section includes: · Eight-line analysis section 'specifies the edge shape of the building and 5 ^ 2 based on the analysis of the photo = Quantitative analysis of the tongue during the mapping process. Specific mapping of the large plane: ^ After the overlapped state, the step becomes extremely steep by the overlapped state. 1 2 · — Collecting electronic data of various kinds of data and using '^' to collect the three photography departments for generating buildings, using electronics: 'Including: Obtaining photographic parameters to obtain photos of the building; the building; The actual attack: Secondly, in order to make the photographic position of the photo lost and the actual landscape and the photographic parameters suitable for the photo are placed in the J data storage section, electronic data and photographic parameters are stored in relation to the photographic position's data. The application is referred to as the patent collection device # 2 of the patent scope, among which, 2061-4944-PF (N); ahddub.ptd p.37 ^ 65810 / 、 Shenjing patent range photography department coefficient camera; photography: the camera is built in the camera 'acquiring the camera- < " r / / λ v — * Make the computer realize the following functions: It can be read by a computer, and the functions of the building can be entered in an electronic data table ^; photos and plane shapes can be entered on the map in the map / 1 = location data and specific; functions of location parameters; universal and According to the input data, the photography of the image angle is to arrange the function of the virtual shape of the plane shape of the building in the quasi-space, and the function of the virtual piece of the electronic data; The shape of the building in the height direction is consistent with the photo as a kind of recording medium. The computer will be used to support the construction of the building's objects-the second-generation electronic data generated by the brain-thinking anti-threat computer. The following function ^ Formula 5 has been recorded on a computer and can be read by a computer, so that the function can be used to input a photo of the building expressed in electronic data and a plane shape. The photographic position of the photo and the building can be viewed. The function of the building that is actually suitable for the required photographic parameters; u The photo on page 38 2061-4944-PF (N); ahddub.ptd 565810 6. The scope of patent application is generating the three yuan In the virtual space of electronic data, the plane shape is arranged according to the relative positional relationship based on the photographing position, and the function of displaying the image of the direction of the building from the photographing position is displayed; The function of projecting the photo into the displayed image according to the position specified by the photographic parameter; and displaying the shape of the building defined by mapping the planar shape in the height direction together in the display Functions within the displayed image. 2061-4944-PF (N); ahddub.ptd p. 39